Paul Saffo moderator - Institute for the Future
Angela Wilkinson - James Martin Institute for Science and Industry at Oxford
Peter Schwartz - Chairman GBN
jim Goodnight - CEO SAS institute.
Ian Goldin - director james martin 21st century school at Oxford.

interesting session. best bits are the rules which seemed to emerge.

Paul started off with a presentation of the panel
Ian - James MArtin endowed a centre to think about the challenges of the next decades....climate change, pandemics, population change, and the like. he focuses on longer term rather than short term. he looks a population dynamics. he also noted that one must be careful about the range of values that one considers. he also noted that the tools that we have can barely deal with the complexity of the current and emerging issues. the do quite a bit of work with nano and bio tech. nano is by its nature very unpredictable. the levels of knowledge in bio, specifically, really make forecasting an interesting new area. it also brings up very fascinating moral issues. technology will potentially drive us to an increased level of predictability as we also conftont all the social issues.

Jim . SAS makes software to help predict. there is a serious skills shortage in this area. the forcasting tools that they use are more around parts or credit card fraud or who to cross-sell to based on information data. they due sales forecasting based on past and future data.

PS - GBN tried to help others make better decisions in the face of uncertainty. he makes multiple scenarios. thus, he will be somewhat right and wrong. the real challenge is to get decision makers to listen so that they can reframe the questions. the problem is not the future part, but the decision map. most decision makers have clear mental maps that have worked for them, and is most likely the wrong one to use to deal with the change that is occurring. ie the real directional impacts of climate change. the future of biology.....what is the future of synthetic biology - it is both transformational and evolutionary.

AW - the past 50 years has seen a proliferation of tools, but we have a poverty of understanding of what works. she believes that forecasting works in some situations, but can also be fatal. what is a strategic decision making capacity?

PS - RULE 1 it is important to know when not to make a forecast.

IG - forecasting tools are really excellent when used appropriately. ie part needs for boeing, or package flows for DHL.

PS - in financial markets there are now very powerful mathematical tools to predict market dynamics over the short term. this works, except when it doesn't. this means that the equations which are used have boundary conditions. when the conditions change, then the equations can, and probably will, fail. another good example was the long-term viability of the sub-prime mortgage market. the challenge is to recognize, if not even to define, what the conditions are.

AW - there is a paradox in forecasting. we are addicted to predictions, yet know that they will be wrong. forecasting and undermine forecasting. forecasting assumes that there is a bit of continuity. when we ask around why they are using forecasting....'because everyone else is using them'.

GEOFF MOORE - 'most of us in the room are making bets every day. i see a disconnect between the scenarios and the mathmatical modelling. why is this and how can they converge? are we on the edge of qualitative forecasting?'

JG - we see forecasting that is really near-term estimations.

IG - we also believe that forecasting can really help - we are working on pandemics, to try to figure out what to do, to deal with the impact. i would rather have the govt based their reactions on his work than the gut reactions. ie who should get

PS - there are two classes of problems -one which has clear boundary conditions and the other which does not. ie a manufacturing line vs society. this means, ther eis a high dependence on the certainty/uncertaintly of the underlying assumptions. every model is a representation of a system.

Rule - any forecasts are better than no forecasts; and bad forecasts are also very useful.

AW - IPCC forecasting or not? she states that there was no double-loop learning. we are really bad a putting complexity tools to use.

Joe Nye - the danger is really that the more explicit one begins to believe in the predction. the vital part is to look at the What ifs.......the forecasting is a baseline. we have to continually challenge the believers.

PS - totally right. it is important to continually challenge their thinking. many prefer the delusion of certainty rather than stating the difficult questions. this leads to trying to understand the weak signals.

AW - each organization have cognitive filters. it is important that organizations have the capacity to catch the weak signals.

PS - Peter what was your worst forecasts?
PS - when i was in Mexico talking about the future of Mexico. we were totally wrong. the lesson was that we did not have enough diversity of opinion in the room.

RULE - never breathe your own exhaust
RULE - understand the mental map of those with whom you are working. if you do not understand their mental map, then you cannot help move them. ie what are the real goals? what is the language they understand? what external information will change the internal mental map?
RULE - forecasting is most successful in well bounded conditions

GEOFF MOORE - three different levels of certainty seem to be talked about here
well bounded - forecasting tools to be used
complexity that is not deterministic, but familiar -
complexity, that is unknown unknown that can only be understood by narrative.

AW - social messyness is an area of interest to me. we also have to respect social complexity. how do we get more effective social; cultural theorys to help.....we are looking at the story of wha tthe future of the city might be.
PG - the future isn't what it used to be. from paul valerie from 1932. the future was predictable until the creation of modern science. we need to understand motivations better.